Why Political Polls Are A "Crock" - An Update To An Earlier Version


As we near the election (16 days away now) the political polls are somewhat useless. One day Hillary is up by 6%, the next day Trump is up by 2%, then Hillary is back on top, and the seesaw continues.

Just take a look at two different projections: 1) Clinton winning large and Trump winning big. One may easily question that the sets of data being used by these two seperate polls are not even close to being from the same planet.

To me, all polling does is try and influenece a mind set which can/will be transferred to others (we usually call this Propaganda.). Tom Robbins offered us a quote on this subject: "One has not only an ability to perceive the world but an ability to alter one's perception of it; more simply, one can change things by the manner in which one looks at them." As an example, take a look at some history from the UK during WWII.

All polling is tabulated using modeling software. And, what's the most applicable "moto" in software? (Besides "I'm almost finished.", I mean.) Answer: Garbage In, Garbage Out! What data is put into the modeling determines what comes out.

As I mentioned in the original version of this article (see below), polls mean very little. The results depend on who is asking the questions, how many are polled, what the questions are, whether the pollees answer the questions honestly, and the demographics of the pollees. We both know that if I ask the right questions to the right set of people I can make the polls say what I want them to say. Or, as the old saying goes: "Figures lie and liars figure." For example, There's This, of course.

There are too many uncontrollable variables to allow for any valid, or semi-valid, prognostications.

Is there any way to discern that the current polling results we see in the news are in any way accurate? The short answer is "NO!". Even if one considers the past accuracy of a polling organization, the variables - especially in this election - are too many and too unverifiable.

An example: With the seesaw nature of the polling data being presented by the media, Bill O'Reilly, of FOX News, set out to do his own, independent polling. He had University graduate students call 250 people from each party's voter rolls in all 50 states. That's 25,000 people - far more than normal polling. What did his poll show? 86% were for Trump and 14% were for Clinton. NOTICE: O'Reilly's poll ONLY contacted Democrat and Republican registered voters - NOT Independents. So, right off the bat, as straightforward O'Reilly's poll seems to have been, a full 42% of registered voters were not contacted.

How accurate was this result? Who knows! But, his results were a far cry from the main media sources' outputs. The same variables mentioned above plague any polling no matter who is doing it and how honest and unbiased (or biased) the polling organization happens to be.

But, I have noticed a somewhat peculiar phenomenon. The sizes of the crowds that show up for the two candidates' rallies do NOT jibe with the polling results the media offers us. Take a look at the below pictures.

One could reasonably assume that if the polling results seem to seesaw back and forth between the two candidates, the rally crowds would, at least to some degree, be on par with the polling results. This is clearly not the case.

Now, I realize that many people don't have the desire to spend an afternoon or evening surrounded by a bunch of people feeling the same way they do about a candidate. It's just too much trouble. But, apparently, many, many do - especially for Trump.

Is this because he is a lot more entertaining than Hillary? Possibly. But, the vast differences in the rally crowd numbers do pose a question.

Are the sizes of the rally crowds more indicative of the real sentiments of the voting public than are the polling data?

I guess we'll just have to wait and see.

ORIGINAL VERSION of this Article:

As we should all be aware by now, the results of political polling should not be trusted for too very much. It all depends on the questions and how they are asked.

Case In Point -

Last evening I received a phone call from a political polling organization. This poll has a specific focus: a "head-to-head" match between Hillary Clinton and Donald Trump for the presidency.

So, I decided to participate in the polling.

By listening to the questions one is able, fairly easily, to determine which political party is sponsoring the polling.

Here are two examples of the questions asked:

1. "To the following statement, please answer with one of the following responses:

      a. Fully Support

      b. Somewhat Support

      c. Undecided

      c. Do Not Support

QUESTION: Do you support Hillary Clinton over Donald Trump for the presidency?"

Somewhat straightforward, wouldn't you say? A fair question, stated clearly.

Now, here's a sample of the questions where Trump was the focus:

1. "To the following statement, please answer with one of the following responses:

      a. Have serious concerns

      b. Have some concerns

      c. Have little concern

      c. Have no concerns

QUESTION: Do you have concerns over Donald Trump's racist attitudes and approaches to undocumented workers, Hispanics, and Muslims here in or entering the United States"?

NOTE: I am not making these questions up. They are exactly as asked!!

This last question was akin to ole "goody", "Do you still beat your wife?" No matter what you answer, it's the wrong answer. The poll gave no possibility of the person polled responding that he/she did not think Trump's remarks were racist.

NOTE: This was not an isolated example. All of the polling questions about Trump were framed in this manner. None of the questions relating to Clinton had a similar bias.

I can just see the results of the polling now:

"62% of those polled have little or no concerns with Donald Trumps' racist remarks."

Now, this poll just happened to be biased against Trump. One could just as easily exist biased against Clinton.

So, when one hears the results of polling, take them with a "grain (or shaker full) of salt".